留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

船舶强框架序贯代理模型辅助遗传优化方法

汪俊泽 王元 易家祥 韩涛 江璞玉 吴嘉蒙 程远胜 刘均

汪俊泽, 王元, 易家祥, 等. 船舶强框架序贯代理模型辅助遗传优化方法[J]. 中国舰船研究, 2021, 16(4): 44–52 doi: 10.19693/j.issn.1673-3185.02118
引用本文: 汪俊泽, 王元, 易家祥, 等. 船舶强框架序贯代理模型辅助遗传优化方法[J]. 中国舰船研究, 2021, 16(4): 44–52 doi: 10.19693/j.issn.1673-3185.02118
WANG J Z, WANG Y, YI J X, et al. Genetic optimization method of ship strong frame based on sequential surrogate model[J]. Chinese Journal of Ship Research, 2021, 16(4): 44–52 doi: 10.19693/j.issn.1673-3185.02118
Citation: WANG J Z, WANG Y, YI J X, et al. Genetic optimization method of ship strong frame based on sequential surrogate model[J]. Chinese Journal of Ship Research, 2021, 16(4): 44–52 doi: 10.19693/j.issn.1673-3185.02118

船舶强框架序贯代理模型辅助遗传优化方法

doi: 10.19693/j.issn.1673-3185.02118
详细信息
    作者简介:

    汪俊泽,男,1997年生,硕士生。研究方向:舰船结构优化设计。E-mail:wang_junze@hust.edu.cn

    吴嘉蒙,男,1976年生,硕士,研究员

    程远胜,男,1962年生,博士,教授,博士生导师。研究方向:结构分析与轻量化设计,结构冲击动力学与防护设计,基于代理模型的优化方法。E-mail:yscheng@hust.edu.cn

    刘均,男,1981年生,博士,副教授,博士生导师。研究方向:结构分析与优化,结构抗爆抗冲击,结构振动与控制。E-mail:hustlj@hust.edu.cn

    通信作者:

    刘均

  • 中图分类号: U663.2

Genetic optimization method of ship strong frame based on sequential surrogate model

  • 摘要:   目的  共同结构规范(CSR)要求下的船舶强框架结构优化存在约束条件多、计算耗时、可行性判断复杂的特点。应用静态代理模型辅助优化算法求解该问题时,因其关注的是模型的整体预测精度,故在样本容量较小的情况下无法保证关键区域的模型预测精度。针对该问题,提出基于序贯代理模型辅助遗传算法的强框架优化方法。  方法  首先,分析CSR对强框架结构的约束要求,根据约束类型,将原始的675条约束缩减为2条积极约束,再对目标函数和约束函数建立代理模型。然后,基于可行性准则,利用遗传算法对代理模型进行优化求解,得到优化解后,计算优化解的真实响应并更新代理模型,再利用期望可行性函数(EFF)准则更新约束代理模型,提高代理模型在约束边界上的精度,如此迭代求解多次,最终得到满足约束条件的全局最优解。  结果  强框架优化结果显示,所提序贯代理模型算法能够在相同,甚至更少的计算资源下得到优于基于静态代理模型优化算法的优化解,最终实现设计区域减重达15.55%。  结论  提出的序贯代理模型算法显著优于静态代理模型算法,在复杂约束下的船舶结构优化问题上有着较好的应用价值。
  • 图  1  强框架结构

    Figure  1.  Strong frame structure

    图  2  强框架中下部分区域设计变量定义示意图

    Figure  2.  Schematic diagram of design variables definition in the middle-lower part of strong frame

    图  3  强框架方案可行性的判断逻辑

    Figure  3.  Judgement logic about infeasible scheme of strong frame

    图  4  基于序贯代理模型辅助遗传算法的强框架优化框架

    Figure  4.  Optimization framework of strong frame based on sequential surrogate assisted genetic algorithm

    图  5  基于静态代理模型辅助遗传算法的强框架优化方法

    Figure  5.  Optimization method of strong frame based on static surrogate assisted genetic algorithm

    图  6  强框架中下区域优化收敛曲线

    Figure  6.  Optimization convergence curves of the middle-lower part of strong frame

    表  1  强框架中下部分区域设计变量及其取值范围

    Table  1.   Definition and ranges of design variables in the middle-lower part of strong frame

    变量取值范围
    区域1竖桁腹板厚$ {x}_{1} $/mm[10.25,14.25]
    区域2竖桁腹板厚$ {x}_{2} $/mm[10.25,14.25]
    区域3竖桁腹板厚$ {x}_{3} $/mm[10.25,14.25]
    区域4竖桁腹板厚$ {x}_{4} $/mm[10.25,14.25]
    区域5肋板厚$ {x}_{5/{\rm{m}}{\rm{m}}} $[10.50,14.50]
    区域6肋板厚$ {x}_{6}/{\rm{m}}{\rm{m}} $[10.50,14.50]
    区域7肋板厚$ {x}_{7}/{\rm{m}}{\rm{m}} $[10.50,14.50]
    区域8肋板厚$ {x}_{8} $/mm[10.50,14.50]
    区域9肋板厚$ {x}_{9}/{\rm{m}}{\rm{m}} $[10.50,14.50]
    区域10肘板厚$ {x}_{10/{\rm{m}}{\rm{m}}} $[10.25,14.25]
    区域11肘板趾端厚$ {x}_{11}/{\rm{m}}{\rm{m}} $[13.25,17.25]
    区域12肘板厚$ {x}_{12} $/mm[10.25,14.25]
    区域13肘板趾端厚$ {x}_{13}/{\rm{m}}{\rm{m}} $[13.25,17.25]
    A面板厚$ {x}_{14} $/mm[17.25,21.25]
    B面板厚$ {x}_{15} $/mm[17.25,21.25]
    下载: 导出CSV

    表  2  强框架中下区域优化结果

    Table  2.   Optimization results of the middle-lower part of strong frame

    初始质量静态代理模型序贯代理模型
    70个样本点100个样本点150个样本点(30+2×20=70个样本点)(60+2×20=100个样本点)
    中下区域总质量/t64.9064.768 764.537 664.89663.903 963.890 7
    总质量下降百分比/%0.006 21.534 81.555 2
    设计区域质量/t6.489 46.358 16.127 06.485 45.493 35.480 1
    设计区域质量下降百分比/%0.061 615.349 615.553 1
    粗网格模型约束满足程度/%−18.55−3.21−3.57−18.74−17.64−12.54
    细网格模型约束满足程度/%−11.262.632.16−3.84−0.85−0.60
    下载: 导出CSV

    表  3  强框架中下区域优化后设计变量取值

    Table  3.   Optimized design variable values of the middle-lower part of strong frame

    变量原始值规范下限静态代理模型
    (150个样本点)
    序贯代理模型
    (30+2×20=70个样本点)(60+2×20=100个样本点)
    区域1竖桁腹板厚$ {x}_{1} $/mm15.7510.2514.2513.7513.75
    区域2竖桁腹板厚$ {x}_{2} $/mm12.7510.2514.2512.2511.75
    区域3竖桁腹板厚$ {x}_{3} $/mm12.7510.2512.7514.2513.25
    区域4竖桁腹板厚$ {x}_{4} $/mm11.7510.2511.7510.2510.25
    区域5肋板厚$ {x}_{5} $/mm12.510.514.5010.5010.50
    区域6肋板厚$ {x}_{6} $/mm12.510.514.5010.5010.50
    区域7肋板厚$ {x}_{7} $/mm11.510.512.5010.5010.50
    区域8肋板厚$ {x}_{8} $/mm11.510.510.5010.5010.50
    区域9肋板厚$ {x}_{9} $/mm12.510.511.5010.5010.50
    区域10肘板厚$ {x}_{10} $/mm11.7510.2513.2510.2510.25
    区域11肘板趾端厚$ {x}_{11} $/mm19.2510.2515.7513.2513.25
    区域12肘板厚$ {x}_{12} $/mm14.2510.2511.2510.2510.25
    区域13肘板趾端厚$ {x}_{13} $/mm19.2510.2517.2517.2513.25
    A面板厚$ {x}_{14} $/mm23.2510.7520.2517.2521.25
    B面板厚$ {x}_{15} $/mm23.2510.7520.7517.2517.25
    下载: 导出CSV
  • [1] 赵留平, 詹大为, 程远胜, 等. 船舶结构优化设计技术研究进展[J]. 中国舰船研究, 2014, 9(4): 1–10. doi: 10.3969/j.issn.1673-3185.2014.04.001

    ZHAO L P, ZHAN D W, CHENG Y S, et al. Review on optimum design methods of ship structures[J]. Chinese Journal of Ship Research, 2014, 9(4): 1–10 (in Chinese). doi: 10.3969/j.issn.1673-3185.2014.04.001
    [2] 郑少平, 陈静, 程远胜, 等. 代理模型技术及其在船舶板架强度和稳定性计算中的应用[J]. 中国造船, 2013, 54(1): 40–51. doi: 10.3969/j.issn.1000-4882.2013.01.007

    ZHENG S P, CHEN J, CHENG Y S, et al. Surrogate models and their application in calculation of strength and stability of ship grillage[J]. Shipbuilding of China, 2013, 54(1): 40–51 (in Chinese). doi: 10.3969/j.issn.1000-4882.2013.01.007
    [3] 夏志, 刘均, 程远胜. 基于代理模型的水下结构物基座阻抗特性快速预报[J]. 中国舰船研究, 2020, 15(3): 81–87.

    XIA Z, LIU J, CHENG Y S. Fast prediction of mechanical impedance of an underwater foundation based on surrogate models[J]. Chinese Journal of Ship Research, 2020, 15(3): 81–87 (in Chinese).
    [4] 陈静, 詹大为, 刘江鹓, 等. 基于Kriging模型的船舶典型双层底板架强度和稳定性全局敏度分析[J]. 中国舰船研究, 2014, 9(1): 72–80. doi: 10.3969/j.issn.1673-3185.2014.01.011

    CHEN J, ZHAN D W, LIU J W, et al. Kriging model-based global sensitivity analysis of strength and stability for ship grillages[J]. Chinese Journal of Ship Research, 2014, 9(1): 72–80 (in Chinese). doi: 10.3969/j.issn.1673-3185.2014.01.011
    [5] KROMANIS R, KRIPAKARAN P. Support vector regression for anomaly detection from measurement histories[J]. Advanced Engineering Informatics, 2013, 27(4): 486–495. doi: 10.1016/j.aei.2013.03.002
    [6] 张干锋, 王德禹. 基于分区间泛化Kriging近似模型的船舶局部结构形状优化[J]. 海洋工程, 2018, 36(6): 46–57.

    ZHANG G F, WANG D Y. Shape optimization of ships' local structure with the partition and generalization Kriging approximation model[J]. The Ocean Engineering, 2018, 36(6): 46–57 (in Chinese).
    [7] 张磊, 胡震. 基于克里金模型的潜水器耐压舱结构优化[J]. 船舶力学, 2020, 24(1): 108–117. doi: 10.3969/j.issn.1007-7294.2020.01.012

    ZHANG L, HU Z. Structural optimization of submersible spherical pressure hull based on Kriging model[J]. Journal of Ship Mechanics, 2020, 24(1): 108–117 (in Chinese). doi: 10.3969/j.issn.1007-7294.2020.01.012
    [8] 龙周, 陈松坤, 王德禹. 基于SMOTE算法的船舶结构可靠性优化设计[J]. 上海交通大学学报, 2019, 53(1): 26–34.

    LONG Z, CHEN S K, WANG D Y. Reliability-based design optimization of ship structures based on SMOTE algorithm[J]. Journal of Shanghai Jiao Tong University, 2019, 53(1): 26–34 (in Chinese).
    [9] JONES D R, SCHONLAU M, WELCH W J. Efficient global optimization of expensive black-box functions[J]. Journal of Global Optimization, 1998, 13(4): 455–492. doi: 10.1023/A:1008306431147
    [10] FORRESTER A I J, KEANE A J. Recent advances in surrogate-based optimization[J]. Progress in Aerospace Sciences, 2009, 45(1/2/3): 50–79.
    [11] GUTMANN H M. A radial basis function method for global optimization[J]. Journal of Global Optimization, 2001, 19(3): 201–227. doi: 10.1023/A:1011255519438
    [12] ZHAN D W, XING H L. Expected improvement for expensive optimization: a review[J]. Journal of Global Optimization, 2020, 78(3): 507–544. doi: 10.1007/s10898-020-00923-x
    [13] YI J X, CHENG Y S, LIU J. An adaptive constraint-handling approach for optimization problems with expensive objective and constraints[C]//Proceedings of 2020 IEEE Congress on Evolutionary Computation. Glasgow, United Kingdom: IEEE, 2020: 1–8.
    [14] BICHON B J, ELDRED M S, SWILER L P, et al. Efficient global reliability analysis for nonlinear implicit performance functions[J]. AIAA Journal, 2008, 46(10): 2459–2468. doi: 10.2514/1.34321
    [15] ECHARD B, GAYTON N, LEMAIRE M. AK-MCS: an active learning reliability method combining Kriging and Monte Carlo simulation[J]. Structural Safety, 2011, 33(2): 145–154. doi: 10.1016/j.strusafe.2011.01.002
    [16] IACS. Common structural rules for bulk carriers and oil tankers[S]. London: International Association of Classification Society, 2019.
    [17] FORRESTER A I J, SÓBESTER A, KEANE A J. Engineering design via surrogate modelling[M]. Southern Gate: John Wiley & Sons Ltd., 2008.
    [18] DEB K. An efficient constraint handling method for genetic algorithms[J]. Computer Methods in Applied Mechanics and Engineering, 2000, 186(2/3/4): 311–338.
  • 加载中
图(6) / 表(3)
计量
  • 文章访问数:  186
  • HTML全文浏览量:  61
  • PDF下载量:  38
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-09-20
  • 修回日期:  2021-02-28
  • 网络出版日期:  2021-06-21
  • 刊出日期:  2021-08-10

目录

    /

    返回文章
    返回